Test your scientific literacy
September 4, 2003 11:03 PM   Subscribe

Test your scientific literacy. 'Do you think you know what science is? You may be surprised.'
posted by plep (22 comments total)
 
I have a Master's Degree... in Science!
posted by Stan Chin at 11:28 PM on September 4, 2003


(via infidels.org).
posted by plep at 11:35 PM on September 4, 2003


/me blinds Stan Chin with science.
posted by mr_crash_davis at 11:35 PM on September 4, 2003


Just reading the stories of the proto-scientists and their wacky explanations for all things was kind of amusing...
An interesting excercise is to try to debunk the old theories using just the empirical knowledge available at the time. It makes you appreciate the non-arbitrariness of these theories...
posted by talos at 1:40 AM on September 5, 2003


a number of the provided answers beg the question, sadly
posted by shoos at 3:20 AM on September 5, 2003


plep: it's ON infidels.org. Perhaps the via in this case is overkill. Thanks for the link though. Enjoyed it.
posted by walrus at 4:38 AM on September 5, 2003


I don't understand the first question ... Anybody?
posted by ZenMasterThis at 4:51 AM on September 5, 2003


walrus: Also 'via' infidels.org, because that was the page which ultimately led me to the link.
:)
posted by plep at 5:37 AM on September 5, 2003


Isn't question 2 and 10 plain wrong, considering that a mathematical theorem can be proven?
And I'd say that anyone that consciously manipulates his own experiments or design them in a way that allows manipulation (question 9) is not a scientist in the sense of the word used here. Perhaps some disciplines, like anthropology, are exceptions but then the question should not have been stated in that way.

ZenMasterThis:
Experiments are not usually carried out in a "I wonder what happens if I mix X and Y?" frame of mind - more like "if hypothesis A is true, mixing X and Y should produce result Z."

Does that clarify it?
posted by spazzm at 6:33 AM on September 5, 2003


Isn't question 2 and 10 plain wrong, considering that a mathematical theorem can be proven?

spazzm: I don't believe that mathematics is normally classified as a science, in the same way that physics or biology are. Mathematics has a very long pedigree, going back to the ancient world; science is a much more recent phenomenon. A mathematical theorem is not the same as a scientific law. (Check out the 'Concluding Observations' section on the page for the author's opinion on the difference between science and other rational fields such as mathematics, philosophy and history).

ZenMasterThis: What spazzm said.
posted by plep at 6:45 AM on September 5, 2003


ZenMasterThis: The author also states the case under the heading 'Experiments are a Goal-Oriented Form of Scientific Observation'.
posted by plep at 6:52 AM on September 5, 2003


whether maths + science are the same or not is a philosophical debate with a long history (search for synthetic and analytic truth).

question 9 is controversial because "manipulate" is a "bad" word that in popular usage implies fakery. that's not what's being implied here. the idea is that experiments are not neutral - they're designed for a prticular task and the results are processed with a certain aim in mind. so they involve a lot of assumptions on the part of the experimenter,

the thing that got me on the test was the clear distinction they make between law and theory. i'm not sure it's that clear cut - if "law" implies causal connection (rather than mere observed correlation) then it's on the same philosophically shaky ground as "theory".
posted by andrew cooke at 7:09 AM on September 5, 2003


But are you making love to him now, crash?
posted by trondant at 7:14 AM on September 5, 2003


I had a philosophy of science class in college. One of the main points of the class is that a scientific theory isn't Truth. It is just a good explanation of the facts at the time, and is subject to change when further facts appear. For example, Newton's laws turn out to be only approximations, eventually replaced by Einstein's relativity theories.
posted by Xoc at 7:28 AM on September 5, 2003


Xoc, if you read this article you'll find that they say Newton's laws weren't replaced, but rather explained by Einstein's theories (it's an example used in the text).
posted by andrew cooke at 8:52 AM on September 5, 2003


Looking at this, I wondered if you couldn't argue that scientific observations remain as fact -- that is, no matter the hypothesis, the gathered data (including the associated bounds of error) is always "true".
posted by weston at 12:54 PM on September 5, 2003


hmmm. i think you could argue that f=ma is still "kind of true" - you could argue that special relativity refines what "m" is, while general relativity unifies that m with the m in g=GMm/r^2. the fight over SR wasn't between newtonian and relativistic physics so much as between relative and absolute frames of reference (and even with relative frames of reference you still have acceleration and inertia). as far as i remember, when i was taught SR we didn't throw away newton, but rather ask what would happen if c were constant in all frames.

about modern laws - maybe there are examples from, say, fractal geometry? maybe you can characterize self-organising systems in some way? i don't know of an example, but it seems like the kind of thing that would be consistent with the way "law" is used in the article - a formal relationship between observables that some theory (which we may not yet have) would explain (ie weston's facts).
posted by andrew cooke at 1:31 PM on September 5, 2003


Re "law": in my experience, there are two ways that word is used in science (and in engineering, which I'm more familiar with). One usage, the one that the quiz's author isn't using, is just that a law is a particularly respected theory. The other usage is that a law is an observed correlation. In this case "law" is used in the sense of "rule" as in straightedge or rule-of-thumb, meaning a pattern or regularity. In this second sense it doesn't indicate a controlling or authoritative influence. "Regular" is etymologically related to "rule", "ruler", etc., which also means "king, authority, lawgiver", etc. --- these two concepts have been commingled in our language for a long time, but they are distinct concepts. It's a subtle distinction at first, but an important one, and it's the one that I suspect the quiz's author is trying to emphasize.

Some examples. "Ohm's law" in electronics is a relation between voltage and current. It's extremely heavily used in electronic design and analysis. However, not all devices follow Ohm's law. This doesn't invalidate Ohm's law; all it means is that you can't look to Ohm's law for a prediction as to how the non-ohmic devices will behave.

Other things might follow a "square law" or "power law", which means that one parameter is related to another by its square or other power. Zipf's Law is an example of a power law, one which relates the length of a word to its frequency of occurrence. Again, no theory is shattered if something is observed not to follow Zipf's Law. But if something is observed to follow Zipf's Law, then you can use that information later, or you can try to develop a theory which is consistent with that result. Any time you see a pattern in the data, it's a potential insight.

Bode's Law describes the spacing of the planets' orbits. It doesn't work very well for the outermost planets and there's no reason to believe that other planetary systems will follow the same pattern. It's just chance. Still, it is a pattern, and in that sense can reasonably be called a "law".
posted by hattifattener at 1:42 PM on September 5, 2003


Mmmmm.... it's all logical positivy.

In my uninformed opinion, the wording in the whole quiz was troubling. The author defines terms, particularly "laws" and "theories" so finely as to be useless in real life. In addition, many of the questions were worded to imply the opposite of what the author appears to have meant.

Q7. "To be scientific one must conduct experiments". While it is true that an individual can make a career out of generating theories (and many do), eventually someone always has to test them with real data. The author reduces to the absurd his "answer".

Q9: The phrase "manipulating experiments" implies fraud to me. While it's true that one tests things with a range of answers in mind, one generally does not try to guess the specific result in advance. For example, when measuring temperature, on expects a number rather than an answer like "a rabbit". This question needs to be rephrased to have more than a trivial meaning.

Q14 "Theories are just ideas about how stuff works." The author loads the question with the word "just", then goes off on an irrelevant emotional harangue about belief. Hypotheses are "just" ideas about how stuff works. Theories are "proven" hypotheses, hence a subset of the class of "ideas of how stuff works" (to get all set theoretical). The author needs to get over his bad self.

Q15 "Laws are more than theories." Echoing the previous question, the author assigns an unwarranted and unsupportable special status to Laws. His distinction that "theories infer, laws describe" is one without a difference. All laws, theories, lemmas and others are descriptions, models, of observations. All theories are made by numerical or descriptive correlations of observations. All laws are empirical and statistical, arising from observation. There is no difference save age. Newton's law of gravity (f=Gmm'/r) is a correlation of measurements, with G being a fit parameter. Darwin's Law is a description based on his observation of finches (mainly). The observation I made the other day, that the aqueous solubility of certain ethers is related to their structure in certain empirically-fitted ways is as much a theory as Newton's or Darwin's "laws". While my crude little model is in no way as useful or nearly as well tested as those other two, it has the same level of significance: it is a descriptive model of the way the world works in its own little domain, based on a correltation between presumed cause and supposed effect. The distinction between a "law" and a "theory" is one of vocabulary. There is no qualitative difference between them.
posted by bonehead at 2:11 PM on September 5, 2003


Weston: That's an interesting question: are "facts" true?

Assuming you discard solipsism, I see it really as a question of trusting the measurement process. Do the constant random perturbations that the experimenter cannot control overwhelm the "facts"? The answer lies in statistics. One tries to tease signal out of the noise, but there is no absolute certainty. Is one using the best statistical model? Are the cause (the variable) and the effect (the result) independent, or do the two depend on something else? These are all interesting questions. You do the best you can, of course, but ultimately you could be fooling yourself.
posted by bonehead at 2:19 PM on September 5, 2003


bonehead - how do you explain the situation where you have one set of observations, which gives a (single) law, and two or more competing theories? it seems to me that this is the essence of the difference between law and theory in the article.

in more detail: you may infer several different laws from the observations, if the observations range over many variables, but if two laws have an equivalent mathematical form (relate the same set of variables in the same way) then they are identical. yet it is possible (normal) to have several competing theories that all fit the observations and which are considered to be distinct (because they make differeing predictions about situations not yet observed). a theory is an explanation, which entails predictions, while a law is an observation, which does not.

[i tend to agree with your views, but having no good reply to this argument.]
posted by andrew cooke at 3:45 PM on September 5, 2003


I still do not understand the special status you assign to a law over a theory. You start with a pool of data. You systematize the data, summarize it, model it with a conceptual framework, be it a mathematical, logical or verbal description, and you have a hypothesis. You assume the null hypothesis to test your model with the data. Assuming this works, you have a theory.

Now you say that you have two competing correlations. The normal thing to do, is to imagine a set of conditions where the theories would predict different results and perform the test. The observation that light bent around the moon during a lunar eclipse separated GR from Newtonian mechanics, for example. If you cannot distinguish the theories given your data, and this is a problem with many current sub-quark "quantum gravity" models, then you have to accept them as equally likely.

I don't buy your statement that "a law is an observation", sir. Indeed, I do not even understand it. An observation, to me, means a measurement, not a conceptual framework for describing observations. We must be very clear here: the framework is not the world. "Laws" are human constructs, not reality.

In the common usage, to me, a "natural law" is simply a very heavily tested theory that is generally applicable, and has some elegance. A "law" is just a really good model or conceptual framework for describing the world, but it is not the world. Ceci n'est pas une pipe.
posted by bonehead at 8:30 PM on September 5, 2003


« Older Getting warmer...   |   George Rarey's War Cartoon Journals Newer »


This thread has been archived and is closed to new comments